I recently got in a mildly heated, shall we say, discussion, on this very topic on the Hi-End Scanning newslist. The person who started the discussion did so on the premise that his sole criteria for image quality was image detail, so going with that qualification, I indulged. His statement that, comparing a scanned 35mm slide or neg with a full frame 24X36mm digital capture, was that there was no possible way that a "mere" 48mb file could equal the optical resolution of a 240mb 8000 ppi drum scanned piece of film with, as he put it, 80 mp of "optical" data. Unfortunately, it was easily shown that a 1DsMKII using sharp lenses at optimum apertures, locked down on a heavy tripod using mirror up cable release, could not only equal film resolution, but easily outperform. Now, granted, he had a lot at stake, as he is in the business of selling drum scanners, he could only believe his outdated theories and not what everyone's eyes clearly saw. The complete absense of grain and noise in the digital captures more than offset the higher resolution of the scan and the smaller grain particles (as compared to pixel site dimension) in the scanned film. To be as fair as possible, and actually having no precoceived notions as to what the real outcome might reveal, I shot several types of fine grained color transparency film and T-Max 100 black and white. I also used the same scanner model and resolution as he had used to make his claims.
In addition to shooting an outdoor real world target that included ever decreasing detail that went beyond the limit of camera/lens/film resolution capture, an image that was criticized by some as not being scientific - something that often happens when someone doesn't like the results, I also shot a resolution test target, in which the digital file was still higher resolution, if not quite as smooth on the radial pie charts.
The outdoor direct sun images showed that the digital captures clearly had 2-3 more stops of effective dymanic range, and possible more if one was making small reproductions where shadow noise wouldn't be a problem. Color neg does have more dynamic range at the expense of fairly large grain and very limited resolving power.
All this is to emphasize that there are many factors that influence image quality, many of which favor modern digital captures and a few of which favor a slightly more organic look of film. It all depends on your intentions.